Water, water everywhere — but where is the science supporting its management in the Delta?

Much of the California public suffering from yet another round of flooding would be astonished to know that the water-export pumps in the south Sacramento-San Joaquin Delta serving agricultural and urban water users in the southern half of the state were operating at less than half of their capacity. State Water Project pumps that can operate at levels up to 8,500 cfs, had been reduced to just 2,000 cfs at the start of January, less than a quarter of their capacity. During the first week of this new year more than 200,000 acre-feet of water a day was flowing into the Delta and then out to San Francisco Bay. Last week that outflow had increased to over 300,000 acre-feet per day. For context, 300,000 acre-feet is about a third of the capacity of Folsom Lake.

As California struggles to recover from three years of intensive drought and as water users in the San Joaquin Valley desperately yearn to restore groundwater supplies, this ongoing management of the state’s increasingly scarce water resources is mystifying. Additionally, the State has a commitment to “co-equal goals” in managing the Delta water supplies — providing a more reliable water supply for California and protecting, restoring, and enhancing the Delta ecosystem. It seems that hapless application of science to water management is much to blame for the state’s inability to meet its co-equal goals and for the Delta’s prolonged recovery.

Current regulations limiting water exports are designed to protect the imperiled delta smelt from losses due to entrainment at Central Valley Project and State Water Project pumps in the south Delta. But those regulations were derived from an assumption that is not supported by the best available scientific information. The assumption behind the regulations is that delta smelt disperse from their habitats in the Suisun Bay and Suisun Marsh following the first big inflow event of the year. That “first flush” through the estuary triggers dispersion of the tiny fish into fresher and more turbid water to spawn, exposing a small percentage of them to potential mortality at the export pumps.

The premise behind constraining water exports during the first two weeks following a first flush is that delta smelt, then on the move, are particularly susceptible to entrainment at the project pumps. Temporarily reducing exports is supposed to limit losses of delta smelt because by springtime delta smelt will have settled in spawning areas well away from the effects of the pumps.

Based on that premise, the U.S. Fish and Wildlife Service (USFWS) and California Department of Fish and Wildlife (CDFW) have imposed restrictions on pump operations during the first big storm or storms of the winter. Those wildlife agencies relied on a study by Polansky and his colleagues that concluded that “the majority of regional movement from juvenile and subadult rearing locations to spawning areas has already happened by the time the [Spring Kodiak Trawl] survey is conducted, that spawning habitat locations are relatively constant within and between years, and that no substantial further restructuring of the population at regional scales occurs afterwards.” (That spring survey starts in mid-January.)  The Polansky team found that the first flush through the Delta does not immediately trigger dispersion by delta smelt, which would expose the population to losses at the water export pumps in the south Delta.

Whether or not the Polansky conclusion is accurate, USFWS and CDFW ostensibly relied on it to institute an Early Winter Pulse Protection action with the requirement “Between December 1 and January 31 each year Permittee shall reduce south Delta exports for 14 consecutive days to maintain a 14-day average OMR index no more negative than -2,000 cfs.” But that reliance is not reconcilable with the study’s conclusion. If the general distribution of delta smelt is not going to change in January, there is no need for a restriction on water exports.

Support for the first flush action is even more tenuous when considering historical data on delta smelt and winter-time outflow through the Delta.  During periods of very high outflow, delta smelt are found mostly in the calmer waters of the north Delta, Suisun Bay, and Suisun Marsh, presumably holding established positions in their habitat and resisting strong westward flows. Accordingly, the majority of delta smelt are not located near the water project pumps between December and the end of January. Although delta smelt historically have been lost to entrainment at the pumps during that period, in the past decade the losses have been very few. The loss of a single delta smelt or several in January is not a sign of high risk to the population or an immediately impending entrainment event. Historical survey data confirm the basic behavior of the delta smelt — in years of very high outflow the vast majority of the delta smelt are located, and stay located, in the northern estuary many miles from the project pumps until at least the end of January.  Increased losses of delta smelt still may occur, but not until later in the season.

USFWS and CDFW developed and implemented restrictions based on a premise regarding the distribution of delta smelt that is not consistent with the government’s own data. Those data confirm that delta smelt in January in years with especially high through-Delta flows are mostly located in the northern portion of the estuary, far from the export pumps. Maps of the distribution of delta smelt from 2006, a year with particularly high outflow, illustrate this clearly.

These maps show the monthly distribution of delta smelt during the water year 2006, a year with a very big first flush, like this year. The solid blue dots show the distribution of delta smelt in the autumn before the first flush. Pink pie segments represent females; blue pie segments represent males. The bigger the pie, the greater the number of delta smelt observed at a location. The maps indicate that delta smelt are not located near the water-project pumps until mid-February in years with large first flush events and that restricting pumping in January does not help delta smelt. Importantly, note that delta smelt are very rarely located near the pumps.  Source: Hamilton Resource Economics, compiled using maps from the California Department of Fish and Wildlife.

In years with especially high outflow, the small fraction of the delta smelt population that appears to move eastward into areas that could be influenced by the export pumps do so many weeks after the onset of the first flush. Data from three of the wettest years that followed the federal listing of delta smelt in 1993 and prior to the initiation of flow-management actions to protect the fish from entrainment in 2008 show that the state’s 14-day prescription for constrained export volumes is inevitably invoked when the majority of delta smelt are nowhere near the pumps. 

Recorded occurrences of delta smelt at the export pumps in years with large first-flush events. The table depicts years of very high inflow into the Delta between 1993, when fish identification protocols were upgraded, and 2007, before implementation of flow management actions to protect delta smelt.  These data indicate that delta smelt movements immediately after first-flush events are not necessarily to spawning areas, that restrictions on export volumes after a first flush are unnecessary to protect delta smelt, and that subsequent redispersion of some delta smelt occurs at least in some wet years.  

In developing regulations to restrict pumping during periods of extreme outflow, the wildlife agencies have either not used or have misused historical data concerning the movement of delta smelt and the timing of those movements. They then have generated water management policy claiming that science supports the policy. But the Polansky investigation of delta smelt population responses to environmental factors was not designed to assess the influence of the first flush on the distribution of delta smelt. It did not consider the influence of inflow events on the fish’s distribution, which changes from season to season and with environmental events, which vary with water years. Nor did the study consider the distribution of delta smelt prior to the beginning of the calendar year; that would have helped determine how static the fish’s distribution may be after a first flush.

The state and federal wildlife agencies have been collecting detailed data on the Delta’s fishes and the environmental factors and conditions that affect them for more than 40 years. But when it comes to developing regulations, analysis of those data appears haphazard. Historical data indicate that given the very high inflows into the Delta over this past month or more, delta smelt have and still are distributed far from the pumps and will not enter the southern Delta at least until February. Given the current hydrodynamic conditions, the regulation curtailing water exports is unnecessary and provides no additional protection to the imperiled delta smelt.

By the time the current regulation expires, California may have lost as much as 200,000 acre-feet of water that could have contributed to water storage in central and southern California – without any impact on the imperiled delta smelt or the other protected fish species that reside in or pass through the Delta. For those who might want to monetize the water that could have been exported, at $300 an acre-foot, 60 million dollars of freshwater went into the bay with no conservation benefits for the Delta’s flagship species.

Achievement of the Delta’s co-equal goals requires discerning guidance from the best available scientific information by discerning resource managers. That has not happened over the past month. Wrong-headed management decisions have done nothing to help the imperiled delta smelt and contributed to perpetuating the economic and social pain caused by years of drought across much of California, from Silicon Valley to the Central Valley and southern California.